ABSTRACT
The type and pattern of immune cell infiltrate in breast cancer is of growing clinical importance as they associate
with response to therapy and are the specific target of immunotherapy. `Cold' cancers that lack infiltrating T cells
exhibit pronounced transforming growth factor ß (TGFß) activity and predict poor outcomes in breast cancer
patients. However, the factors that influence the genesis of the type of tumor immune microenvironment (TiME)
have yet to be defined. We found that radiation-preceded breast cancers in women treated with radiation therapy
for Hodgkin's lymphoma are significantly enriched for TiME devoid of lymphocytes and rich in myeloid cells,
TGFß and cyclooxygenase 2. We used a Trp53 null mammary chimera model to determine the factors
underpinning of this unexpected difference. Tumors with an immunosuppressive TiME lacking lymphocytes
arose only in irradiated mice, even when the transplant was not irradiated, indicating host biology was key, as
well as in mice lacking functional adaptive immunity, pointing to a role for innate immunity. Strikingly, transient
aspirin treatment before cancer developed blocked the development of cold tumors. We hypothesize that
systemic inflammation provokes the development of tumors with immunosuppressive, cold TiME. Chronic low-
level inflammation from aging, obesity, stress and chronic syndromes following viral infection is common. Here
we will test the specific hypothesis that inflammation-induced TGFß during carcinogenesis alters tissue-resident
myeloid cells to promote the genesis of cancers with an immunosuppressive TiME. AIM 1 will use state-of-the-
art analysis of cytokines and immune characteristics that correlate with the development of tumors with cold
TiME using a novel biobank of blood, plasma, bone marrow, spleen, and nonmalignant mammary glands and
their associated cancers as a function of inflammation or anti-inflammatory aspirin conditions at 4-, 8- and 18-
months post-treatment. The relevance of these findings will be tested by immunoprofiling women with breast
cancer. AIM 2 will use parabiosis to test whether factors circulating during systemic inflammation contribute and
use macrophage depletion and a mouse in which myeloid cells cannot signal through TGFß to test whether
circulating TGFß elicits monocyte activation to promote the development of cold TiME. AIM 3 will analyze the
resulting high-content data using deep learning and bioinformatics methods to identify tumor subtypes and to
infer key events. The main goal of our study is to test the innovative hypothesis that inflammation-induced TGFß
promotes cold tumors by altering tissue-resident myeloid cells during carcinogenesis. Our proposal to conduct
systematic, high content analysis and modeling of the mechanisms by which breast cancers develop with an
immunosuppressive TiME is highly significant in view of the growing clinical importance of the TiME.